Validation and methodology.

A platform that computes clinical dose metrics owes you its methodology. Radial's is public lineage, refusal over guesswork, and validation against known answers.

A DVH engine with public lineage

Radial's dose-volume computation descends from dicompyler-core, the open-source engine this team has developed and maintained for over a decade, cited and cross-checked by the community long before Radial existed. The platform's results are benchmarked against it, which is why "how were the DVHs computed?" in your methods section becomes one sentence.

It refuses to guess

Real archives contain plans whose records are incomplete or ambiguous (e.g., doses that may already be summed, intervals that would have to be inferred). Where the data doesn't support an honest number, Radial says so instead of computing past it. A blank with a reason is a better foundation than a plausible guess.

Deterministic where it matters

Curation decisions are evidence-based and reproducible: the same structure against the same dictionary yields the same proposal, and every confirmation is recorded. There is no opaque model between your archive and your numbers.

Validated against known answers

Radial's metrics are checked against reference cases with known-answer values: fixtures the engine must reproduce exactly, verified on every change and benchmarked against dicompyler-core. It is how the numbers stay honest internally, and the same fixtures anchor a design partner's own validation.